1 .. SPDX-License-Identifier: GPL-2.0 2 3 ========================================= 4 Overview of the Linux Virtual File System 5 ========================================= 6 7 Original author: Richard Gooch <rgooch@atnf.csiro.au> 8 9 - Copyright (C) 1999 Richard Gooch 10 - Copyright (C) 2005 Pekka Enberg 11 12 13 Introduction 14 ============ 15 16 The Virtual File System (also known as the Virtual Filesystem Switch) is 17 the software layer in the kernel that provides the filesystem interface 18 to userspace programs. It also provides an abstraction within the 19 kernel which allows different filesystem implementations to coexist. 20 21 VFS system calls open(2), stat(2), read(2), write(2), chmod(2) and so on 22 are called from a process context. Filesystem locking is described in 23 the document Documentation/filesystems/locking.rst. 24 25 26 Directory Entry Cache (dcache) 27 ------------------------------ 28 29 The VFS implements the open(2), stat(2), chmod(2), and similar system 30 calls. The pathname argument that is passed to them is used by the VFS 31 to search through the directory entry cache (also known as the dentry 32 cache or dcache). This provides a very fast look-up mechanism to 33 translate a pathname (filename) into a specific dentry. Dentries live 34 in RAM and are never saved to disc: they exist only for performance. 35 36 The dentry cache is meant to be a view into your entire filespace. As 37 most computers cannot fit all dentries in the RAM at the same time, some 38 bits of the cache are missing. In order to resolve your pathname into a 39 dentry, the VFS may have to resort to creating dentries along the way, 40 and then loading the inode. This is done by looking up the inode. 41 42 43 The Inode Object 44 ---------------- 45 46 An individual dentry usually has a pointer to an inode. Inodes are 47 filesystem objects such as regular files, directories, FIFOs and other 48 beasts. They live either on the disc (for block device filesystems) or 49 in the memory (for pseudo filesystems). Inodes that live on the disc 50 are copied into the memory when required and changes to the inode are 51 written back to disc. A single inode can be pointed to by multiple 52 dentries (hard links, for example, do this). 53 54 To look up an inode requires that the VFS calls the lookup() method of 55 the parent directory inode. This method is installed by the specific 56 filesystem implementation that the inode lives in. Once the VFS has the 57 required dentry (and hence the inode), we can do all those boring things 58 like open(2) the file, or stat(2) it to peek at the inode data. The 59 stat(2) operation is fairly simple: once the VFS has the dentry, it 60 peeks at the inode data and passes some of it back to userspace. 61 62 63 The File Object 64 --------------- 65 66 Opening a file requires another operation: allocation of a file 67 structure (this is the kernel-side implementation of file descriptors). 68 The freshly allocated file structure is initialized with a pointer to 69 the dentry and a set of file operation member functions. These are 70 taken from the inode data. The open() file method is then called so the 71 specific filesystem implementation can do its work. You can see that 72 this is another switch performed by the VFS. The file structure is 73 placed into the file descriptor table for the process. 74 75 Reading, writing and closing files (and other assorted VFS operations) 76 is done by using the userspace file descriptor to grab the appropriate 77 file structure, and then calling the required file structure method to 78 do whatever is required. For as long as the file is open, it keeps the 79 dentry in use, which in turn means that the VFS inode is still in use. 80 81 82 Registering and Mounting a Filesystem 83 ===================================== 84 85 To register and unregister a filesystem, use the following API 86 functions: 87 88 .. code-block:: c 89 90 #include <linux/fs.h> 91 92 extern int register_filesystem(struct file_system_type *); 93 extern int unregister_filesystem(struct file_system_type *); 94 95 The passed struct file_system_type describes your filesystem. When a 96 request is made to mount a filesystem onto a directory in your 97 namespace, the VFS will call the appropriate mount() method for the 98 specific filesystem. New vfsmount referring to the tree returned by 99 ->mount() will be attached to the mountpoint, so that when pathname 100 resolution reaches the mountpoint it will jump into the root of that 101 vfsmount. 102 103 You can see all filesystems that are registered to the kernel in the 104 file /proc/filesystems. 105 106 107 struct file_system_type 108 ----------------------- 109 110 This describes the filesystem. As of kernel 2.6.39, the following 111 members are defined: 112 113 .. code-block:: c 114 115 struct file_system_operations { 116 const char *name; 117 int fs_flags; 118 struct dentry *(*mount) (struct file_system_type *, int, 119 const char *, void *); 120 void (*kill_sb) (struct super_block *); 121 struct module *owner; 122 struct file_system_type * next; 123 struct list_head fs_supers; 124 struct lock_class_key s_lock_key; 125 struct lock_class_key s_umount_key; 126 }; 127 128 ``name`` 129 the name of the filesystem type, such as "ext2", "iso9660", 130 "msdos" and so on 131 132 ``fs_flags`` 133 various flags (i.e. FS_REQUIRES_DEV, FS_NO_DCACHE, etc.) 134 135 ``mount`` 136 the method to call when a new instance of this filesystem should 137 be mounted 138 139 ``kill_sb`` 140 the method to call when an instance of this filesystem should be 141 shut down 142 143 144 ``owner`` 145 for internal VFS use: you should initialize this to THIS_MODULE 146 in most cases. 147 148 ``next`` 149 for internal VFS use: you should initialize this to NULL 150 151 s_lock_key, s_umount_key: lockdep-specific 152 153 The mount() method has the following arguments: 154 155 ``struct file_system_type *fs_type`` 156 describes the filesystem, partly initialized by the specific 157 filesystem code 158 159 ``int flags`` 160 mount flags 161 162 ``const char *dev_name`` 163 the device name we are mounting. 164 165 ``void *data`` 166 arbitrary mount options, usually comes as an ASCII string (see 167 "Mount Options" section) 168 169 The mount() method must return the root dentry of the tree requested by 170 caller. An active reference to its superblock must be grabbed and the 171 superblock must be locked. On failure it should return ERR_PTR(error). 172 173 The arguments match those of mount(2) and their interpretation depends 174 on filesystem type. E.g. for block filesystems, dev_name is interpreted 175 as block device name, that device is opened and if it contains a 176 suitable filesystem image the method creates and initializes struct 177 super_block accordingly, returning its root dentry to caller. 178 179 ->mount() may choose to return a subtree of existing filesystem - it 180 doesn't have to create a new one. The main result from the caller's 181 point of view is a reference to dentry at the root of (sub)tree to be 182 attached; creation of new superblock is a common side effect. 183 184 The most interesting member of the superblock structure that the mount() 185 method fills in is the "s_op" field. This is a pointer to a "struct 186 super_operations" which describes the next level of the filesystem 187 implementation. 188 189 Usually, a filesystem uses one of the generic mount() implementations 190 and provides a fill_super() callback instead. The generic variants are: 191 192 ``mount_bdev`` 193 mount a filesystem residing on a block device 194 195 ``mount_nodev`` 196 mount a filesystem that is not backed by a device 197 198 ``mount_single`` 199 mount a filesystem which shares the instance between all mounts 200 201 A fill_super() callback implementation has the following arguments: 202 203 ``struct super_block *sb`` 204 the superblock structure. The callback must initialize this 205 properly. 206 207 ``void *data`` 208 arbitrary mount options, usually comes as an ASCII string (see 209 "Mount Options" section) 210 211 ``int silent`` 212 whether or not to be silent on error 213 214 215 The Superblock Object 216 ===================== 217 218 A superblock object represents a mounted filesystem. 219 220 221 struct super_operations 222 ----------------------- 223 224 This describes how the VFS can manipulate the superblock of your 225 filesystem. As of kernel 2.6.22, the following members are defined: 226 227 .. code-block:: c 228 229 struct super_operations { 230 struct inode *(*alloc_inode)(struct super_block *sb); 231 void (*destroy_inode)(struct inode *); 232 233 void (*dirty_inode) (struct inode *, int flags); 234 int (*write_inode) (struct inode *, int); 235 void (*drop_inode) (struct inode *); 236 void (*delete_inode) (struct inode *); 237 void (*put_super) (struct super_block *); 238 int (*sync_fs)(struct super_block *sb, int wait); 239 int (*freeze_fs) (struct super_block *); 240 int (*unfreeze_fs) (struct super_block *); 241 int (*statfs) (struct dentry *, struct kstatfs *); 242 int (*remount_fs) (struct super_block *, int *, char *); 243 void (*clear_inode) (struct inode *); 244 void (*umount_begin) (struct super_block *); 245 246 int (*show_options)(struct seq_file *, struct dentry *); 247 248 ssize_t (*quota_read)(struct super_block *, int, char *, size_t, loff_t); 249 ssize_t (*quota_write)(struct super_block *, int, const char *, size_t, loff_t); 250 int (*nr_cached_objects)(struct super_block *); 251 void (*free_cached_objects)(struct super_block *, int); 252 }; 253 254 All methods are called without any locks being held, unless otherwise 255 noted. This means that most methods can block safely. All methods are 256 only called from a process context (i.e. not from an interrupt handler 257 or bottom half). 258 259 ``alloc_inode`` 260 this method is called by alloc_inode() to allocate memory for 261 struct inode and initialize it. If this function is not 262 defined, a simple 'struct inode' is allocated. Normally 263 alloc_inode will be used to allocate a larger structure which 264 contains a 'struct inode' embedded within it. 265 266 ``destroy_inode`` 267 this method is called by destroy_inode() to release resources 268 allocated for struct inode. It is only required if 269 ->alloc_inode was defined and simply undoes anything done by 270 ->alloc_inode. 271 272 ``dirty_inode`` 273 this method is called by the VFS to mark an inode dirty. 274 275 ``write_inode`` 276 this method is called when the VFS needs to write an inode to 277 disc. The second parameter indicates whether the write should 278 be synchronous or not, not all filesystems check this flag. 279 280 ``drop_inode`` 281 called when the last access to the inode is dropped, with the 282 inode->i_lock spinlock held. 283 284 This method should be either NULL (normal UNIX filesystem 285 semantics) or "generic_delete_inode" (for filesystems that do 286 not want to cache inodes - causing "delete_inode" to always be 287 called regardless of the value of i_nlink) 288 289 The "generic_delete_inode()" behavior is equivalent to the old 290 practice of using "force_delete" in the put_inode() case, but 291 does not have the races that the "force_delete()" approach had. 292 293 ``delete_inode`` 294 called when the VFS wants to delete an inode 295 296 ``put_super`` 297 called when the VFS wishes to free the superblock 298 (i.e. unmount). This is called with the superblock lock held 299 300 ``sync_fs`` 301 called when VFS is writing out all dirty data associated with a 302 superblock. The second parameter indicates whether the method 303 should wait until the write out has been completed. Optional. 304 305 ``freeze_fs`` 306 called when VFS is locking a filesystem and forcing it into a 307 consistent state. This method is currently used by the Logical 308 Volume Manager (LVM). 309 310 ``unfreeze_fs`` 311 called when VFS is unlocking a filesystem and making it writable 312 again. 313 314 ``statfs`` 315 called when the VFS needs to get filesystem statistics. 316 317 ``remount_fs`` 318 called when the filesystem is remounted. This is called with 319 the kernel lock held 320 321 ``clear_inode`` 322 called then the VFS clears the inode. Optional 323 324 ``umount_begin`` 325 called when the VFS is unmounting a filesystem. 326 327 ``show_options`` 328 called by the VFS to show mount options for /proc/<pid>/mounts. 329 (see "Mount Options" section) 330 331 ``quota_read`` 332 called by the VFS to read from filesystem quota file. 333 334 ``quota_write`` 335 called by the VFS to write to filesystem quota file. 336 337 ``nr_cached_objects`` 338 called by the sb cache shrinking function for the filesystem to 339 return the number of freeable cached objects it contains. 340 Optional. 341 342 ``free_cache_objects`` 343 called by the sb cache shrinking function for the filesystem to 344 scan the number of objects indicated to try to free them. 345 Optional, but any filesystem implementing this method needs to 346 also implement ->nr_cached_objects for it to be called 347 correctly. 348 349 We can't do anything with any errors that the filesystem might 350 encountered, hence the void return type. This will never be 351 called if the VM is trying to reclaim under GFP_NOFS conditions, 352 hence this method does not need to handle that situation itself. 353 354 Implementations must include conditional reschedule calls inside 355 any scanning loop that is done. This allows the VFS to 356 determine appropriate scan batch sizes without having to worry 357 about whether implementations will cause holdoff problems due to 358 large scan batch sizes. 359 360 Whoever sets up the inode is responsible for filling in the "i_op" 361 field. This is a pointer to a "struct inode_operations" which describes 362 the methods that can be performed on individual inodes. 363 364 365 struct xattr_handlers 366 --------------------- 367 368 On filesystems that support extended attributes (xattrs), the s_xattr 369 superblock field points to a NULL-terminated array of xattr handlers. 370 Extended attributes are name:value pairs. 371 372 ``name`` 373 Indicates that the handler matches attributes with the specified 374 name (such as "system.posix_acl_access"); the prefix field must 375 be NULL. 376 377 ``prefix`` 378 Indicates that the handler matches all attributes with the 379 specified name prefix (such as "user."); the name field must be 380 NULL. 381 382 ``list`` 383 Determine if attributes matching this xattr handler should be 384 listed for a particular dentry. Used by some listxattr 385 implementations like generic_listxattr. 386 387 ``get`` 388 Called by the VFS to get the value of a particular extended 389 attribute. This method is called by the getxattr(2) system 390 call. 391 392 ``set`` 393 Called by the VFS to set the value of a particular extended 394 attribute. When the new value is NULL, called to remove a 395 particular extended attribute. This method is called by the 396 setxattr(2) and removexattr(2) system calls. 397 398 When none of the xattr handlers of a filesystem match the specified 399 attribute name or when a filesystem doesn't support extended attributes, 400 the various ``*xattr(2)`` system calls return -EOPNOTSUPP. 401 402 403 The Inode Object 404 ================ 405 406 An inode object represents an object within the filesystem. 407 408 409 struct inode_operations 410 ----------------------- 411 412 This describes how the VFS can manipulate an inode in your filesystem. 413 As of kernel 2.6.22, the following members are defined: 414 415 .. code-block:: c 416 417 struct inode_operations { 418 int (*create) (struct inode *,struct dentry *, umode_t, bool); 419 struct dentry * (*lookup) (struct inode *,struct dentry *, unsigned int); 420 int (*link) (struct dentry *,struct inode *,struct dentry *); 421 int (*unlink) (struct inode *,struct dentry *); 422 int (*symlink) (struct inode *,struct dentry *,const char *); 423 int (*mkdir) (struct inode *,struct dentry *,umode_t); 424 int (*rmdir) (struct inode *,struct dentry *); 425 int (*mknod) (struct inode *,struct dentry *,umode_t,dev_t); 426 int (*rename) (struct inode *, struct dentry *, 427 struct inode *, struct dentry *, unsigned int); 428 int (*readlink) (struct dentry *, char __user *,int); 429 const char *(*get_link) (struct dentry *, struct inode *, 430 struct delayed_call *); 431 int (*permission) (struct inode *, int); 432 int (*get_acl)(struct inode *, int); 433 int (*setattr) (struct dentry *, struct iattr *); 434 int (*getattr) (const struct path *, struct kstat *, u32, unsigned int); 435 ssize_t (*listxattr) (struct dentry *, char *, size_t); 436 void (*update_time)(struct inode *, struct timespec *, int); 437 int (*atomic_open)(struct inode *, struct dentry *, struct file *, 438 unsigned open_flag, umode_t create_mode); 439 int (*tmpfile) (struct inode *, struct dentry *, umode_t); 440 }; 441 442 Again, all methods are called without any locks being held, unless 443 otherwise noted. 444 445 ``create`` 446 called by the open(2) and creat(2) system calls. Only required 447 if you want to support regular files. The dentry you get should 448 not have an inode (i.e. it should be a negative dentry). Here 449 you will probably call d_instantiate() with the dentry and the 450 newly created inode 451 452 ``lookup`` 453 called when the VFS needs to look up an inode in a parent 454 directory. The name to look for is found in the dentry. This 455 method must call d_add() to insert the found inode into the 456 dentry. The "i_count" field in the inode structure should be 457 incremented. If the named inode does not exist a NULL inode 458 should be inserted into the dentry (this is called a negative 459 dentry). Returning an error code from this routine must only be 460 done on a real error, otherwise creating inodes with system 461 calls like create(2), mknod(2), mkdir(2) and so on will fail. 462 If you wish to overload the dentry methods then you should 463 initialise the "d_dop" field in the dentry; this is a pointer to 464 a struct "dentry_operations". This method is called with the 465 directory inode semaphore held 466 467 ``link`` 468 called by the link(2) system call. Only required if you want to 469 support hard links. You will probably need to call 470 d_instantiate() just as you would in the create() method 471 472 ``unlink`` 473 called by the unlink(2) system call. Only required if you want 474 to support deleting inodes 475 476 ``symlink`` 477 called by the symlink(2) system call. Only required if you want 478 to support symlinks. You will probably need to call 479 d_instantiate() just as you would in the create() method 480 481 ``mkdir`` 482 called by the mkdir(2) system call. Only required if you want 483 to support creating subdirectories. You will probably need to 484 call d_instantiate() just as you would in the create() method 485 486 ``rmdir`` 487 called by the rmdir(2) system call. Only required if you want 488 to support deleting subdirectories 489 490 ``mknod`` 491 called by the mknod(2) system call to create a device (char, 492 block) inode or a named pipe (FIFO) or socket. Only required if 493 you want to support creating these types of inodes. You will 494 probably need to call d_instantiate() just as you would in the 495 create() method 496 497 ``rename`` 498 called by the rename(2) system call to rename the object to have 499 the parent and name given by the second inode and dentry. 500 501 The filesystem must return -EINVAL for any unsupported or 502 unknown flags. Currently the following flags are implemented: 503 (1) RENAME_NOREPLACE: this flag indicates that if the target of 504 the rename exists the rename should fail with -EEXIST instead of 505 replacing the target. The VFS already checks for existence, so 506 for local filesystems the RENAME_NOREPLACE implementation is 507 equivalent to plain rename. 508 (2) RENAME_EXCHANGE: exchange source and target. Both must 509 exist; this is checked by the VFS. Unlike plain rename, source 510 and target may be of different type. 511 512 ``get_link`` 513 called by the VFS to follow a symbolic link to the inode it 514 points to. Only required if you want to support symbolic links. 515 This method returns the symlink body to traverse (and possibly 516 resets the current position with nd_jump_link()). If the body 517 won't go away until the inode is gone, nothing else is needed; 518 if it needs to be otherwise pinned, arrange for its release by 519 having get_link(..., ..., done) do set_delayed_call(done, 520 destructor, argument). In that case destructor(argument) will 521 be called once VFS is done with the body you've returned. May 522 be called in RCU mode; that is indicated by NULL dentry 523 argument. If request can't be handled without leaving RCU mode, 524 have it return ERR_PTR(-ECHILD). 525 526 If the filesystem stores the symlink target in ->i_link, the 527 VFS may use it directly without calling ->get_link(); however, 528 ->get_link() must still be provided. ->i_link must not be 529 freed until after an RCU grace period. Writing to ->i_link 530 post-iget() time requires a 'release' memory barrier. 531 532 ``readlink`` 533 this is now just an override for use by readlink(2) for the 534 cases when ->get_link uses nd_jump_link() or object is not in 535 fact a symlink. Normally filesystems should only implement 536 ->get_link for symlinks and readlink(2) will automatically use 537 that. 538 539 ``permission`` 540 called by the VFS to check for access rights on a POSIX-like 541 filesystem. 542 543 May be called in rcu-walk mode (mask & MAY_NOT_BLOCK). If in 544 rcu-walk mode, the filesystem must check the permission without 545 blocking or storing to the inode. 546 547 If a situation is encountered that rcu-walk cannot handle, 548 return 549 -ECHILD and it will be called again in ref-walk mode. 550 551 ``setattr`` 552 called by the VFS to set attributes for a file. This method is 553 called by chmod(2) and related system calls. 554 555 ``getattr`` 556 called by the VFS to get attributes of a file. This method is 557 called by stat(2) and related system calls. 558 559 ``listxattr`` 560 called by the VFS to list all extended attributes for a given 561 file. This method is called by the listxattr(2) system call. 562 563 ``update_time`` 564 called by the VFS to update a specific time or the i_version of 565 an inode. If this is not defined the VFS will update the inode 566 itself and call mark_inode_dirty_sync. 567 568 ``atomic_open`` 569 called on the last component of an open. Using this optional 570 method the filesystem can look up, possibly create and open the 571 file in one atomic operation. If it wants to leave actual 572 opening to the caller (e.g. if the file turned out to be a 573 symlink, device, or just something filesystem won't do atomic 574 open for), it may signal this by returning finish_no_open(file, 575 dentry). This method is only called if the last component is 576 negative or needs lookup. Cached positive dentries are still 577 handled by f_op->open(). If the file was created, FMODE_CREATED 578 flag should be set in file->f_mode. In case of O_EXCL the 579 method must only succeed if the file didn't exist and hence 580 FMODE_CREATED shall always be set on success. 581 582 ``tmpfile`` 583 called in the end of O_TMPFILE open(). Optional, equivalent to 584 atomically creating, opening and unlinking a file in given 585 directory. 586 587 588 The Address Space Object 589 ======================== 590 591 The address space object is used to group and manage pages in the page 592 cache. It can be used to keep track of the pages in a file (or anything 593 else) and also track the mapping of sections of the file into process 594 address spaces. 595 596 There are a number of distinct yet related services that an 597 address-space can provide. These include communicating memory pressure, 598 page lookup by address, and keeping track of pages tagged as Dirty or 599 Writeback. 600 601 The first can be used independently to the others. The VM can try to 602 either write dirty pages in order to clean them, or release clean pages 603 in order to reuse them. To do this it can call the ->writepage method 604 on dirty pages, and ->releasepage on clean pages with PagePrivate set. 605 Clean pages without PagePrivate and with no external references will be 606 released without notice being given to the address_space. 607 608 To achieve this functionality, pages need to be placed on an LRU with 609 lru_cache_add and mark_page_active needs to be called whenever the page 610 is used. 611 612 Pages are normally kept in a radix tree index by ->index. This tree 613 maintains information about the PG_Dirty and PG_Writeback status of each 614 page, so that pages with either of these flags can be found quickly. 615 616 The Dirty tag is primarily used by mpage_writepages - the default 617 ->writepages method. It uses the tag to find dirty pages to call 618 ->writepage on. If mpage_writepages is not used (i.e. the address 619 provides its own ->writepages) , the PAGECACHE_TAG_DIRTY tag is almost 620 unused. write_inode_now and sync_inode do use it (through 621 __sync_single_inode) to check if ->writepages has been successful in 622 writing out the whole address_space. 623 624 The Writeback tag is used by filemap*wait* and sync_page* functions, via 625 filemap_fdatawait_range, to wait for all writeback to complete. 626 627 An address_space handler may attach extra information to a page, 628 typically using the 'private' field in the 'struct page'. If such 629 information is attached, the PG_Private flag should be set. This will 630 cause various VM routines to make extra calls into the address_space 631 handler to deal with that data. 632 633 An address space acts as an intermediate between storage and 634 application. Data is read into the address space a whole page at a 635 time, and provided to the application either by copying of the page, or 636 by memory-mapping the page. Data is written into the address space by 637 the application, and then written-back to storage typically in whole 638 pages, however the address_space has finer control of write sizes. 639 640 The read process essentially only requires 'readpage'. The write 641 process is more complicated and uses write_begin/write_end or 642 set_page_dirty to write data into the address_space, and writepage and 643 writepages to writeback data to storage. 644 645 Adding and removing pages to/from an address_space is protected by the 646 inode's i_mutex. 647 648 When data is written to a page, the PG_Dirty flag should be set. It 649 typically remains set until writepage asks for it to be written. This 650 should clear PG_Dirty and set PG_Writeback. It can be actually written 651 at any point after PG_Dirty is clear. Once it is known to be safe, 652 PG_Writeback is cleared. 653 654 Writeback makes use of a writeback_control structure to direct the 655 operations. This gives the writepage and writepages operations some 656 information about the nature of and reason for the writeback request, 657 and the constraints under which it is being done. It is also used to 658 return information back to the caller about the result of a writepage or 659 writepages request. 660 661 662 Handling errors during writeback 663 -------------------------------- 664 665 Most applications that do buffered I/O will periodically call a file 666 synchronization call (fsync, fdatasync, msync or sync_file_range) to 667 ensure that data written has made it to the backing store. When there 668 is an error during writeback, they expect that error to be reported when 669 a file sync request is made. After an error has been reported on one 670 request, subsequent requests on the same file descriptor should return 671 0, unless further writeback errors have occurred since the previous file 672 syncronization. 673 674 Ideally, the kernel would report errors only on file descriptions on 675 which writes were done that subsequently failed to be written back. The 676 generic pagecache infrastructure does not track the file descriptions 677 that have dirtied each individual page however, so determining which 678 file descriptors should get back an error is not possible. 679 680 Instead, the generic writeback error tracking infrastructure in the 681 kernel settles for reporting errors to fsync on all file descriptions 682 that were open at the time that the error occurred. In a situation with 683 multiple writers, all of them will get back an error on a subsequent 684 fsync, even if all of the writes done through that particular file 685 descriptor succeeded (or even if there were no writes on that file 686 descriptor at all). 687 688 Filesystems that wish to use this infrastructure should call 689 mapping_set_error to record the error in the address_space when it 690 occurs. Then, after writing back data from the pagecache in their 691 file->fsync operation, they should call file_check_and_advance_wb_err to 692 ensure that the struct file's error cursor has advanced to the correct 693 point in the stream of errors emitted by the backing device(s). 694 695 696 struct address_space_operations 697 ------------------------------- 698 699 This describes how the VFS can manipulate mapping of a file to page 700 cache in your filesystem. The following members are defined: 701 702 .. code-block:: c 703 704 struct address_space_operations { 705 int (*writepage)(struct page *page, struct writeback_control *wbc); 706 int (*readpage)(struct file *, struct page *); 707 int (*writepages)(struct address_space *, struct writeback_control *); 708 int (*set_page_dirty)(struct page *page); 709 void (*readahead)(struct readahead_control *); 710 int (*readpages)(struct file *filp, struct address_space *mapping, 711 struct list_head *pages, unsigned nr_pages); 712 int (*write_begin)(struct file *, struct address_space *mapping, 713 loff_t pos, unsigned len, unsigned flags, 714 struct page **pagep, void **fsdata); 715 int (*write_end)(struct file *, struct address_space *mapping, 716 loff_t pos, unsigned len, unsigned copied, 717 struct page *page, void *fsdata); 718 sector_t (*bmap)(struct address_space *, sector_t); 719 void (*invalidatepage) (struct page *, unsigned int, unsigned int); 720 int (*releasepage) (struct page *, int); 721 void (*freepage)(struct page *); 722 ssize_t (*direct_IO)(struct kiocb *, struct iov_iter *iter); 723 /* isolate a page for migration */ 724 bool (*isolate_page) (struct page *, isolate_mode_t); 725 /* migrate the contents of a page to the specified target */ 726 int (*migratepage) (struct page *, struct page *); 727 /* put migration-failed page back to right list */ 728 void (*putback_page) (struct page *); 729 int (*launder_page) (struct page *); 730 731 int (*is_partially_uptodate) (struct page *, unsigned long, 732 unsigned long); 733 void (*is_dirty_writeback) (struct page *, bool *, bool *); 734 int (*error_remove_page) (struct mapping *mapping, struct page *page); 735 int (*swap_activate)(struct file *); 736 int (*swap_deactivate)(struct file *); 737 }; 738 739 ``writepage`` 740 called by the VM to write a dirty page to backing store. This 741 may happen for data integrity reasons (i.e. 'sync'), or to free 742 up memory (flush). The difference can be seen in 743 wbc->sync_mode. The PG_Dirty flag has been cleared and 744 PageLocked is true. writepage should start writeout, should set 745 PG_Writeback, and should make sure the page is unlocked, either 746 synchronously or asynchronously when the write operation 747 completes. 748 749 If wbc->sync_mode is WB_SYNC_NONE, ->writepage doesn't have to 750 try too hard if there are problems, and may choose to write out 751 other pages from the mapping if that is easier (e.g. due to 752 internal dependencies). If it chooses not to start writeout, it 753 should return AOP_WRITEPAGE_ACTIVATE so that the VM will not 754 keep calling ->writepage on that page. 755 756 See the file "Locking" for more details. 757 758 ``readpage`` 759 called by the VM to read a page from backing store. The page 760 will be Locked when readpage is called, and should be unlocked 761 and marked uptodate once the read completes. If ->readpage 762 discovers that it needs to unlock the page for some reason, it 763 can do so, and then return AOP_TRUNCATED_PAGE. In this case, 764 the page will be relocated, relocked and if that all succeeds, 765 ->readpage will be called again. 766 767 ``writepages`` 768 called by the VM to write out pages associated with the 769 address_space object. If wbc->sync_mode is WB_SYNC_ALL, then 770 the writeback_control will specify a range of pages that must be 771 written out. If it is WB_SYNC_NONE, then a nr_to_write is 772 given and that many pages should be written if possible. If no 773 ->writepages is given, then mpage_writepages is used instead. 774 This will choose pages from the address space that are tagged as 775 DIRTY and will pass them to ->writepage. 776 777 ``set_page_dirty`` 778 called by the VM to set a page dirty. This is particularly 779 needed if an address space attaches private data to a page, and 780 that data needs to be updated when a page is dirtied. This is 781 called, for example, when a memory mapped page gets modified. 782 If defined, it should set the PageDirty flag, and the 783 PAGECACHE_TAG_DIRTY tag in the radix tree. 784 785 ``readahead`` 786 Called by the VM to read pages associated with the address_space 787 object. The pages are consecutive in the page cache and are 788 locked. The implementation should decrement the page refcount 789 after starting I/O on each page. Usually the page will be 790 unlocked by the I/O completion handler. If the filesystem decides 791 to stop attempting I/O before reaching the end of the readahead 792 window, it can simply return. The caller will decrement the page 793 refcount and unlock the remaining pages for you. Set PageUptodate 794 if the I/O completes successfully. Setting PageError on any page 795 will be ignored; simply unlock the page if an I/O error occurs. 796 797 ``readpages`` 798 called by the VM to read pages associated with the address_space 799 object. This is essentially just a vector version of readpage. 800 Instead of just one page, several pages are requested. 801 readpages is only used for read-ahead, so read errors are 802 ignored. If anything goes wrong, feel free to give up. 803 This interface is deprecated and will be removed by the end of 804 2020; implement readahead instead. 805 806 ``write_begin`` 807 Called by the generic buffered write code to ask the filesystem 808 to prepare to write len bytes at the given offset in the file. 809 The address_space should check that the write will be able to 810 complete, by allocating space if necessary and doing any other 811 internal housekeeping. If the write will update parts of any 812 basic-blocks on storage, then those blocks should be pre-read 813 (if they haven't been read already) so that the updated blocks 814 can be written out properly. 815 816 The filesystem must return the locked pagecache page for the 817 specified offset, in ``*pagep``, for the caller to write into. 818 819 It must be able to cope with short writes (where the length 820 passed to write_begin is greater than the number of bytes copied 821 into the page). 822 823 flags is a field for AOP_FLAG_xxx flags, described in 824 include/linux/fs.h. 825 826 A void * may be returned in fsdata, which then gets passed into 827 write_end. 828 829 Returns 0 on success; < 0 on failure (which is the error code), 830 in which case write_end is not called. 831 832 ``write_end`` 833 After a successful write_begin, and data copy, write_end must be 834 called. len is the original len passed to write_begin, and 835 copied is the amount that was able to be copied. 836 837 The filesystem must take care of unlocking the page and 838 releasing it refcount, and updating i_size. 839 840 Returns < 0 on failure, otherwise the number of bytes (<= 841 'copied') that were able to be copied into pagecache. 842 843 ``bmap`` 844 called by the VFS to map a logical block offset within object to 845 physical block number. This method is used by the FIBMAP ioctl 846 and for working with swap-files. To be able to swap to a file, 847 the file must have a stable mapping to a block device. The swap 848 system does not go through the filesystem but instead uses bmap 849 to find out where the blocks in the file are and uses those 850 addresses directly. 851 852 ``invalidatepage`` 853 If a page has PagePrivate set, then invalidatepage will be 854 called when part or all of the page is to be removed from the 855 address space. This generally corresponds to either a 856 truncation, punch hole or a complete invalidation of the address 857 space (in the latter case 'offset' will always be 0 and 'length' 858 will be PAGE_SIZE). Any private data associated with the page 859 should be updated to reflect this truncation. If offset is 0 860 and length is PAGE_SIZE, then the private data should be 861 released, because the page must be able to be completely 862 discarded. This may be done by calling the ->releasepage 863 function, but in this case the release MUST succeed. 864 865 ``releasepage`` 866 releasepage is called on PagePrivate pages to indicate that the 867 page should be freed if possible. ->releasepage should remove 868 any private data from the page and clear the PagePrivate flag. 869 If releasepage() fails for some reason, it must indicate failure 870 with a 0 return value. releasepage() is used in two distinct 871 though related cases. The first is when the VM finds a clean 872 page with no active users and wants to make it a free page. If 873 ->releasepage succeeds, the page will be removed from the 874 address_space and become free. 875 876 The second case is when a request has been made to invalidate 877 some or all pages in an address_space. This can happen through 878 the fadvise(POSIX_FADV_DONTNEED) system call or by the 879 filesystem explicitly requesting it as nfs and 9fs do (when they 880 believe the cache may be out of date with storage) by calling 881 invalidate_inode_pages2(). If the filesystem makes such a call, 882 and needs to be certain that all pages are invalidated, then its 883 releasepage will need to ensure this. Possibly it can clear the 884 PageUptodate bit if it cannot free private data yet. 885 886 ``freepage`` 887 freepage is called once the page is no longer visible in the 888 page cache in order to allow the cleanup of any private data. 889 Since it may be called by the memory reclaimer, it should not 890 assume that the original address_space mapping still exists, and 891 it should not block. 892 893 ``direct_IO`` 894 called by the generic read/write routines to perform direct_IO - 895 that is IO requests which bypass the page cache and transfer 896 data directly between the storage and the application's address 897 space. 898 899 ``isolate_page`` 900 Called by the VM when isolating a movable non-lru page. If page 901 is successfully isolated, VM marks the page as PG_isolated via 902 __SetPageIsolated. 903 904 ``migrate_page`` 905 This is used to compact the physical memory usage. If the VM 906 wants to relocate a page (maybe off a memory card that is 907 signalling imminent failure) it will pass a new page and an old 908 page to this function. migrate_page should transfer any private 909 data across and update any references that it has to the page. 910 911 ``putback_page`` 912 Called by the VM when isolated page's migration fails. 913 914 ``launder_page`` 915 Called before freeing a page - it writes back the dirty page. 916 To prevent redirtying the page, it is kept locked during the 917 whole operation. 918 919 ``is_partially_uptodate`` 920 Called by the VM when reading a file through the pagecache when 921 the underlying blocksize != pagesize. If the required block is 922 up to date then the read can complete without needing the IO to 923 bring the whole page up to date. 924 925 ``is_dirty_writeback`` 926 Called by the VM when attempting to reclaim a page. The VM uses 927 dirty and writeback information to determine if it needs to 928 stall to allow flushers a chance to complete some IO. 929 Ordinarily it can use PageDirty and PageWriteback but some 930 filesystems have more complex state (unstable pages in NFS 931 prevent reclaim) or do not set those flags due to locking 932 problems. This callback allows a filesystem to indicate to the 933 VM if a page should be treated as dirty or writeback for the 934 purposes of stalling. 935 936 ``error_remove_page`` 937 normally set to generic_error_remove_page if truncation is ok 938 for this address space. Used for memory failure handling. 939 Setting this implies you deal with pages going away under you, 940 unless you have them locked or reference counts increased. 941 942 ``swap_activate`` 943 Called when swapon is used on a file to allocate space if 944 necessary and pin the block lookup information in memory. A 945 return value of zero indicates success, in which case this file 946 can be used to back swapspace. 947 948 ``swap_deactivate`` 949 Called during swapoff on files where swap_activate was 950 successful. 951 952 953 The File Object 954 =============== 955 956 A file object represents a file opened by a process. This is also known 957 as an "open file description" in POSIX parlance. 958 959 960 struct file_operations 961 ---------------------- 962 963 This describes how the VFS can manipulate an open file. As of kernel 964 4.18, the following members are defined: 965 966 .. code-block:: c 967 968 struct file_operations { 969 struct module *owner; 970 loff_t (*llseek) (struct file *, loff_t, int); 971 ssize_t (*read) (struct file *, char __user *, size_t, loff_t *); 972 ssize_t (*write) (struct file *, const char __user *, size_t, loff_t *); 973 ssize_t (*read_iter) (struct kiocb *, struct iov_iter *); 974 ssize_t (*write_iter) (struct kiocb *, struct iov_iter *); 975 int (*iopoll)(struct kiocb *kiocb, bool spin); 976 int (*iterate) (struct file *, struct dir_context *); 977 int (*iterate_shared) (struct file *, struct dir_context *); 978 __poll_t (*poll) (struct file *, struct poll_table_struct *); 979 long (*unlocked_ioctl) (struct file *, unsigned int, unsigned long); 980 long (*compat_ioctl) (struct file *, unsigned int, unsigned long); 981 int (*mmap) (struct file *, struct vm_area_struct *); 982 int (*open) (struct inode *, struct file *); 983 int (*flush) (struct file *, fl_owner_t id); 984 int (*release) (struct inode *, struct file *); 985 int (*fsync) (struct file *, loff_t, loff_t, int datasync); 986 int (*fasync) (int, struct file *, int); 987 int (*lock) (struct file *, int, struct file_lock *); 988 ssize_t (*sendpage) (struct file *, struct page *, int, size_t, loff_t *, int); 989 unsigned long (*get_unmapped_area)(struct file *, unsigned long, unsigned long, unsigned long, unsigned long); 990 int (*check_flags)(int); 991 int (*flock) (struct file *, int, struct file_lock *); 992 ssize_t (*splice_write)(struct pipe_inode_info *, struct file *, loff_t *, size_t, unsigned int); 993 ssize_t (*splice_read)(struct file *, loff_t *, struct pipe_inode_info *, size_t, unsigned int); 994 int (*setlease)(struct file *, long, struct file_lock **, void **); 995 long (*fallocate)(struct file *file, int mode, loff_t offset, 996 loff_t len); 997 void (*show_fdinfo)(struct seq_file *m, struct file *f); 998 #ifndef CONFIG_MMU 999 unsigned (*mmap_capabilities)(struct file *); 1000 #endif 1001 ssize_t (*copy_file_range)(struct file *, loff_t, struct file *, loff_t, size_t, unsigned int); 1002 loff_t (*remap_file_range)(struct file *file_in, loff_t pos_in, 1003 struct file *file_out, loff_t pos_out, 1004 loff_t len, unsigned int remap_flags); 1005 int (*fadvise)(struct file *, loff_t, loff_t, int); 1006 }; 1007 1008 Again, all methods are called without any locks being held, unless 1009 otherwise noted. 1010 1011 ``llseek`` 1012 called when the VFS needs to move the file position index 1013 1014 ``read`` 1015 called by read(2) and related system calls 1016 1017 ``read_iter`` 1018 possibly asynchronous read with iov_iter as destination 1019 1020 ``write`` 1021 called by write(2) and related system calls 1022 1023 ``write_iter`` 1024 possibly asynchronous write with iov_iter as source 1025 1026 ``iopoll`` 1027 called when aio wants to poll for completions on HIPRI iocbs 1028 1029 ``iterate`` 1030 called when the VFS needs to read the directory contents 1031 1032 ``iterate_shared`` 1033 called when the VFS needs to read the directory contents when 1034 filesystem supports concurrent dir iterators 1035 1036 ``poll`` 1037 called by the VFS when a process wants to check if there is 1038 activity on this file and (optionally) go to sleep until there 1039 is activity. Called by the select(2) and poll(2) system calls 1040 1041 ``unlocked_ioctl`` 1042 called by the ioctl(2) system call. 1043 1044 ``compat_ioctl`` 1045 called by the ioctl(2) system call when 32 bit system calls are 1046 used on 64 bit kernels. 1047 1048 ``mmap`` 1049 called by the mmap(2) system call 1050 1051 ``open`` 1052 called by the VFS when an inode should be opened. When the VFS 1053 opens a file, it creates a new "struct file". It then calls the 1054 open method for the newly allocated file structure. You might 1055 think that the open method really belongs in "struct 1056 inode_operations", and you may be right. I think it's done the 1057 way it is because it makes filesystems simpler to implement. 1058 The open() method is a good place to initialize the 1059 "private_data" member in the file structure if you want to point 1060 to a device structure 1061 1062 ``flush`` 1063 called by the close(2) system call to flush a file 1064 1065 ``release`` 1066 called when the last reference to an open file is closed 1067 1068 ``fsync`` 1069 called by the fsync(2) system call. Also see the section above 1070 entitled "Handling errors during writeback". 1071 1072 ``fasync`` 1073 called by the fcntl(2) system call when asynchronous 1074 (non-blocking) mode is enabled for a file 1075 1076 ``lock`` 1077 called by the fcntl(2) system call for F_GETLK, F_SETLK, and 1078 F_SETLKW commands 1079 1080 ``get_unmapped_area`` 1081 called by the mmap(2) system call 1082 1083 ``check_flags`` 1084 called by the fcntl(2) system call for F_SETFL command 1085 1086 ``flock`` 1087 called by the flock(2) system call 1088 1089 ``splice_write`` 1090 called by the VFS to splice data from a pipe to a file. This 1091 method is used by the splice(2) system call 1092 1093 ``splice_read`` 1094 called by the VFS to splice data from file to a pipe. This 1095 method is used by the splice(2) system call 1096 1097 ``setlease`` 1098 called by the VFS to set or release a file lock lease. setlease 1099 implementations should call generic_setlease to record or remove 1100 the lease in the inode after setting it. 1101 1102 ``fallocate`` 1103 called by the VFS to preallocate blocks or punch a hole. 1104 1105 ``copy_file_range`` 1106 called by the copy_file_range(2) system call. 1107 1108 ``remap_file_range`` 1109 called by the ioctl(2) system call for FICLONERANGE and FICLONE 1110 and FIDEDUPERANGE commands to remap file ranges. An 1111 implementation should remap len bytes at pos_in of the source 1112 file into the dest file at pos_out. Implementations must handle 1113 callers passing in len == 0; this means "remap to the end of the 1114 source file". The return value should the number of bytes 1115 remapped, or the usual negative error code if errors occurred 1116 before any bytes were remapped. The remap_flags parameter 1117 accepts REMAP_FILE_* flags. If REMAP_FILE_DEDUP is set then the 1118 implementation must only remap if the requested file ranges have 1119 identical contents. If REMAP_FILE_CAN_SHORTEN is set, the caller is 1120 ok with the implementation shortening the request length to 1121 satisfy alignment or EOF requirements (or any other reason). 1122 1123 ``fadvise`` 1124 possibly called by the fadvise64() system call. 1125 1126 Note that the file operations are implemented by the specific 1127 filesystem in which the inode resides. When opening a device node 1128 (character or block special) most filesystems will call special 1129 support routines in the VFS which will locate the required device 1130 driver information. These support routines replace the filesystem file 1131 operations with those for the device driver, and then proceed to call 1132 the new open() method for the file. This is how opening a device file 1133 in the filesystem eventually ends up calling the device driver open() 1134 method. 1135 1136 1137 Directory Entry Cache (dcache) 1138 ============================== 1139 1140 1141 struct dentry_operations 1142 ------------------------ 1143 1144 This describes how a filesystem can overload the standard dentry 1145 operations. Dentries and the dcache are the domain of the VFS and the 1146 individual filesystem implementations. Device drivers have no business 1147 here. These methods may be set to NULL, as they are either optional or 1148 the VFS uses a default. As of kernel 2.6.22, the following members are 1149 defined: 1150 1151 .. code-block:: c 1152 1153 struct dentry_operations { 1154 int (*d_revalidate)(struct dentry *, unsigned int); 1155 int (*d_weak_revalidate)(struct dentry *, unsigned int); 1156 int (*d_hash)(const struct dentry *, struct qstr *); 1157 int (*d_compare)(const struct dentry *, 1158 unsigned int, const char *, const struct qstr *); 1159 int (*d_delete)(const struct dentry *); 1160 int (*d_init)(struct dentry *); 1161 void (*d_release)(struct dentry *); 1162 void (*d_iput)(struct dentry *, struct inode *); 1163 char *(*d_dname)(struct dentry *, char *, int); 1164 struct vfsmount *(*d_automount)(struct path *); 1165 int (*d_manage)(const struct path *, bool); 1166 struct dentry *(*d_real)(struct dentry *, const struct inode *); 1167 }; 1168 1169 ``d_revalidate`` 1170 called when the VFS needs to revalidate a dentry. This is 1171 called whenever a name look-up finds a dentry in the dcache. 1172 Most local filesystems leave this as NULL, because all their 1173 dentries in the dcache are valid. Network filesystems are 1174 different since things can change on the server without the 1175 client necessarily being aware of it. 1176 1177 This function should return a positive value if the dentry is 1178 still valid, and zero or a negative error code if it isn't. 1179 1180 d_revalidate may be called in rcu-walk mode (flags & 1181 LOOKUP_RCU). If in rcu-walk mode, the filesystem must 1182 revalidate the dentry without blocking or storing to the dentry, 1183 d_parent and d_inode should not be used without care (because 1184 they can change and, in d_inode case, even become NULL under 1185 us). 1186 1187 If a situation is encountered that rcu-walk cannot handle, 1188 return 1189 -ECHILD and it will be called again in ref-walk mode. 1190 1191 ``_weak_revalidate`` 1192 called when the VFS needs to revalidate a "jumped" dentry. This 1193 is called when a path-walk ends at dentry that was not acquired 1194 by doing a lookup in the parent directory. This includes "/", 1195 "." and "..", as well as procfs-style symlinks and mountpoint 1196 traversal. 1197 1198 In this case, we are less concerned with whether the dentry is 1199 still fully correct, but rather that the inode is still valid. 1200 As with d_revalidate, most local filesystems will set this to 1201 NULL since their dcache entries are always valid. 1202 1203 This function has the same return code semantics as 1204 d_revalidate. 1205 1206 d_weak_revalidate is only called after leaving rcu-walk mode. 1207 1208 ``d_hash`` 1209 called when the VFS adds a dentry to the hash table. The first 1210 dentry passed to d_hash is the parent directory that the name is 1211 to be hashed into. 1212 1213 Same locking and synchronisation rules as d_compare regarding 1214 what is safe to dereference etc. 1215 1216 ``d_compare`` 1217 called to compare a dentry name with a given name. The first 1218 dentry is the parent of the dentry to be compared, the second is 1219 the child dentry. len and name string are properties of the 1220 dentry to be compared. qstr is the name to compare it with. 1221 1222 Must be constant and idempotent, and should not take locks if 1223 possible, and should not or store into the dentry. Should not 1224 dereference pointers outside the dentry without lots of care 1225 (eg. d_parent, d_inode, d_name should not be used). 1226 1227 However, our vfsmount is pinned, and RCU held, so the dentries 1228 and inodes won't disappear, neither will our sb or filesystem 1229 module. ->d_sb may be used. 1230 1231 It is a tricky calling convention because it needs to be called 1232 under "rcu-walk", ie. without any locks or references on things. 1233 1234 ``d_delete`` 1235 called when the last reference to a dentry is dropped and the 1236 dcache is deciding whether or not to cache it. Return 1 to 1237 delete immediately, or 0 to cache the dentry. Default is NULL 1238 which means to always cache a reachable dentry. d_delete must 1239 be constant and idempotent. 1240 1241 ``d_init`` 1242 called when a dentry is allocated 1243 1244 ``d_release`` 1245 called when a dentry is really deallocated 1246 1247 ``d_iput`` 1248 called when a dentry loses its inode (just prior to its being 1249 deallocated). The default when this is NULL is that the VFS 1250 calls iput(). If you define this method, you must call iput() 1251 yourself 1252 1253 ``d_dname`` 1254 called when the pathname of a dentry should be generated. 1255 Useful for some pseudo filesystems (sockfs, pipefs, ...) to 1256 delay pathname generation. (Instead of doing it when dentry is 1257 created, it's done only when the path is needed.). Real 1258 filesystems probably dont want to use it, because their dentries 1259 are present in global dcache hash, so their hash should be an 1260 invariant. As no lock is held, d_dname() should not try to 1261 modify the dentry itself, unless appropriate SMP safety is used. 1262 CAUTION : d_path() logic is quite tricky. The correct way to 1263 return for example "Hello" is to put it at the end of the 1264 buffer, and returns a pointer to the first char. 1265 dynamic_dname() helper function is provided to take care of 1266 this. 1267 1268 Example : 1269 1270 .. code-block:: c 1271 1272 static char *pipefs_dname(struct dentry *dent, char *buffer, int buflen) 1273 { 1274 return dynamic_dname(dentry, buffer, buflen, "pipe:[%lu]", 1275 dentry->d_inode->i_ino); 1276 } 1277 1278 ``d_automount`` 1279 called when an automount dentry is to be traversed (optional). 1280 This should create a new VFS mount record and return the record 1281 to the caller. The caller is supplied with a path parameter 1282 giving the automount directory to describe the automount target 1283 and the parent VFS mount record to provide inheritable mount 1284 parameters. NULL should be returned if someone else managed to 1285 make the automount first. If the vfsmount creation failed, then 1286 an error code should be returned. If -EISDIR is returned, then 1287 the directory will be treated as an ordinary directory and 1288 returned to pathwalk to continue walking. 1289 1290 If a vfsmount is returned, the caller will attempt to mount it 1291 on the mountpoint and will remove the vfsmount from its 1292 expiration list in the case of failure. The vfsmount should be 1293 returned with 2 refs on it to prevent automatic expiration - the 1294 caller will clean up the additional ref. 1295 1296 This function is only used if DCACHE_NEED_AUTOMOUNT is set on 1297 the dentry. This is set by __d_instantiate() if S_AUTOMOUNT is 1298 set on the inode being added. 1299 1300 ``d_manage`` 1301 called to allow the filesystem to manage the transition from a 1302 dentry (optional). This allows autofs, for example, to hold up 1303 clients waiting to explore behind a 'mountpoint' while letting 1304 the daemon go past and construct the subtree there. 0 should be 1305 returned to let the calling process continue. -EISDIR can be 1306 returned to tell pathwalk to use this directory as an ordinary 1307 directory and to ignore anything mounted on it and not to check 1308 the automount flag. Any other error code will abort pathwalk 1309 completely. 1310 1311 If the 'rcu_walk' parameter is true, then the caller is doing a 1312 pathwalk in RCU-walk mode. Sleeping is not permitted in this 1313 mode, and the caller can be asked to leave it and call again by 1314 returning -ECHILD. -EISDIR may also be returned to tell 1315 pathwalk to ignore d_automount or any mounts. 1316 1317 This function is only used if DCACHE_MANAGE_TRANSIT is set on 1318 the dentry being transited from. 1319 1320 ``d_real`` 1321 overlay/union type filesystems implement this method to return 1322 one of the underlying dentries hidden by the overlay. It is 1323 used in two different modes: 1324 1325 Called from file_dentry() it returns the real dentry matching 1326 the inode argument. The real dentry may be from a lower layer 1327 already copied up, but still referenced from the file. This 1328 mode is selected with a non-NULL inode argument. 1329 1330 With NULL inode the topmost real underlying dentry is returned. 1331 1332 Each dentry has a pointer to its parent dentry, as well as a hash list 1333 of child dentries. Child dentries are basically like files in a 1334 directory. 1335 1336 1337 Directory Entry Cache API 1338 -------------------------- 1339 1340 There are a number of functions defined which permit a filesystem to 1341 manipulate dentries: 1342 1343 ``dget`` 1344 open a new handle for an existing dentry (this just increments 1345 the usage count) 1346 1347 ``dput`` 1348 close a handle for a dentry (decrements the usage count). If 1349 the usage count drops to 0, and the dentry is still in its 1350 parent's hash, the "d_delete" method is called to check whether 1351 it should be cached. If it should not be cached, or if the 1352 dentry is not hashed, it is deleted. Otherwise cached dentries 1353 are put into an LRU list to be reclaimed on memory shortage. 1354 1355 ``d_drop`` 1356 this unhashes a dentry from its parents hash list. A subsequent 1357 call to dput() will deallocate the dentry if its usage count 1358 drops to 0 1359 1360 ``d_delete`` 1361 delete a dentry. If there are no other open references to the 1362 dentry then the dentry is turned into a negative dentry (the 1363 d_iput() method is called). If there are other references, then 1364 d_drop() is called instead 1365 1366 ``d_add`` 1367 add a dentry to its parents hash list and then calls 1368 d_instantiate() 1369 1370 ``d_instantiate`` 1371 add a dentry to the alias hash list for the inode and updates 1372 the "d_inode" member. The "i_count" member in the inode 1373 structure should be set/incremented. If the inode pointer is 1374 NULL, the dentry is called a "negative dentry". This function 1375 is commonly called when an inode is created for an existing 1376 negative dentry 1377 1378 ``d_lookup`` 1379 look up a dentry given its parent and path name component It 1380 looks up the child of that given name from the dcache hash 1381 table. If it is found, the reference count is incremented and 1382 the dentry is returned. The caller must use dput() to free the 1383 dentry when it finishes using it. 1384 1385 1386 Mount Options 1387 ============= 1388 1389 1390 Parsing options 1391 --------------- 1392 1393 On mount and remount the filesystem is passed a string containing a 1394 comma separated list of mount options. The options can have either of 1395 these forms: 1396 1397 option 1398 option=value 1399 1400 The <linux/parser.h> header defines an API that helps parse these 1401 options. There are plenty of examples on how to use it in existing 1402 filesystems. 1403 1404 1405 Showing options 1406 --------------- 1407 1408 If a filesystem accepts mount options, it must define show_options() to 1409 show all the currently active options. The rules are: 1410 1411 - options MUST be shown which are not default or their values differ 1412 from the default 1413 1414 - options MAY be shown which are enabled by default or have their 1415 default value 1416 1417 Options used only internally between a mount helper and the kernel (such 1418 as file descriptors), or which only have an effect during the mounting 1419 (such as ones controlling the creation of a journal) are exempt from the 1420 above rules. 1421 1422 The underlying reason for the above rules is to make sure, that a mount 1423 can be accurately replicated (e.g. umounting and mounting again) based 1424 on the information found in /proc/mounts. 1425 1426 1427 Resources 1428 ========= 1429 1430 (Note some of these resources are not up-to-date with the latest kernel 1431 version.) 1432 1433 Creating Linux virtual filesystems. 2002 1434 <https://lwn.net/Articles/13325/> 1435 1436 The Linux Virtual File-system Layer by Neil Brown. 1999 1437 <http://www.cse.unsw.edu.au/~neilb/oss/linux-commentary/vfs.html> 1438 1439 A tour of the Linux VFS by Michael K. Johnson. 1996 1440 <https://www.tldp.org/LDP/khg/HyperNews/get/fs/vfstour.html> 1441 1442 A small trail through the Linux kernel by Andries Brouwer. 2001 1443 <https://www.win.tue.nl/~aeb/linux/vfs/trail.html> 1444